Alzheimer's disease (AD) likely begins several decades before the onset of clinical symptoms, and preclinical changes such as hippocampal atrophy and early cognitive decline are strong predictors of dementia risk later in life. In addition, oxidative stress is thought to be an important, early pathogenic event in AD, thus antioxidants could be highly relevant to the study of hippocampal volume and early cognitive decrements. Experimental studies show that antioxidants can protect against neuronal degeneration, although epidemiologic studies of antioxidant supplements have found no substantial cognitive benefit. Nonetheless, preliminary dietary studies are intriguing, suggesting that the wider variety of antioxidants available in foods may be most important for reducing brain aging. Thus, the proposed research will explore the effects of dietary intake of antioxidant nutrients, antioxidant-rich foods, and total antioxidant capacity (assessed by a diet-based "antioxidant score") on hippocampal atrophy and early cognitive decline. These analyses will be conducted in two, large, prospective cohorts of older adults;specifically, analyses of hippocampal atrophy (using magnetic resonance imaging to estimate hippocampal volume) will be carried out in the Rotterdam Scan Study, and those of early cognitive decline (based on a validated telephone battery of cognitive assessments) in the Nurses'Health Study. Dietary information in both cohorts is collected using semi-quantitative food frequency questionnaires validated in their respective populations. To quantify antioxidant relationships with these outcomes, multivariate linear regression and random effects models will be used and consideration given to covariate adjustments for a variety of health and lifestyle factors. Together, these results could eventually help to inform preventive strategies for dementia, as intervention probably will be most effective at the earliest likely stages of disease.